Home / s0mbrad / opentcm_discovery · GitHub

open-tcm-discovery SKILL.md skillB

open-tcm-discovery from s0mbrad/opentcm_discovery is an agent skill: a SKILL.md folder an AI agent loads when a task calls for it.

Install it, pinned
mdr add s0mbrad/opentcm_discovery/open-tcm-discovery@git:20261018.34f43f3

Pins git:20261018.34f43f3 in mdr.lock. Pin the fingerprint instead.

Audit grade
B, 16 of 17 checks passed
Kind
SKILL.md skill
Repository
s0mbrad/opentcm_discovery
Path
SKILL.md
Versions recorded
5
Latest version
git:20261018.34f43f3, committed Oct 18, 2026
First version recorded
Oct 13, 2026
Size
42,336 bytes
Repository stars
31
License
MIT

Unified TCM drug-discovery skill: disease-first screening (disease to KEGG/SymMap pathways, candidate herbs, OB-filtered ingredients, hypergeometric FDR), cross-database evidence screening (PubChem, ChEMBL, SymMap, STRING, Open Targets, Pharos), PPI network and hub analysis, dual-engine enrichment consensus, Wikidata/UniProt ID bridging, molecular library building with QC and dedup, AutoDock Vina docking, contact and H-bond analysis, geometric PPI modeling, PyMOL publication figures, runtime…

Collected from public GitHub. Every version is saved exactly as it was and never changes. Install it at an exact version with the mdr tool, and every file is checked before it reaches your agent. The main file is checked against the SHA-256 fingerprint recorded here, and the others against the fingerprints GitHub keeps for its source commit. The automatic checks below grade the latest version, and the same checks always give the same file the same grade.

What the file says

# 中药小分子检索与对接

这个技能检索中药小分子，并按步骤做筛选和对接。先从疾病检索，再做多库核对、网络和富集、分子库，然后做对接和论文图。一次运行结束之后，可以做自迭代。

服务里注册了 81 个工具：检索分析 27 个，对接建模 8 个，智能层 8 个，并行层 4 个，证据与出图 4 个，P1 有 3 个，P2 有 1 个，依赖自举 1 个，数据整理 6 个，TCMSP 旧站检索 1 个，成分名对齐 1 个，靶点表读入 1 个，方剂克重 1 个，对接复现门槛 1 个，HERB 详情 1 个，名称候选 1 个，公开查询 11 个，本地特化 1 个。另外有 20 个脚本、11 步本地 DAG、15 份参考文档。工具名录见 `docs/TOOLS.md`。

智能层负责出计划，并执行在线步骤。抓取药材成分要 30–60 分钟，用 `scripts/disease_pathway_workflow.py` 另跑。`depth=full` 时，引擎只把这条命令登记为延后步骤 D3，不在同一次请求里抓取。

## 先做这几步

1. **看总览**。调用 `unified_overview`，看模块、工具和依赖是否正常，并读本文件的约定。
2. **出计划**。调用 `tcm_smart_plan`。可以传 goal，也可以传 disease、compounds、genes。计划里带决策点 D1–D6。
3. **按深度执行**。调用 `tcm_smart_run(goal=...)`。在线链是：M1 通路；D1 分支之后做 M3 网络或富集；M2 多库核对；P1 ADMET；P0 文献共现；P2 临床试验。文献口袋和 Vina 都齐了，才进入 M5 对接。`depth=full` 不在这次调用里跑四步药材脚本，只把命令写入延后步骤 D3。缺文献口袋残基或 Vina 时暂停，并写出续跑参数。不要猜口袋。用同一个 `session_id` 续跑时，已完成的步骤不重跑。Vina 按这个顺序找：调用参数，然后 `VINA_EXE`，然后 `bin/vina.exe`，最后 PATH。
4. **比较多个候选**。`tcm_multi_path_run(compounds=[...])` 让每个候选各跑一条链（共识、ADME、可选对接），再出横向对比和综合排名。一个受体对多个配体，用 `dock_batch_screening`：并行 Vina，先做 ADME 预过滤，再排名。看网络里哪个节点脆，用 `tcm_network_simulation`：knockout 扫描、蒙特卡洛对照、z-score。
…

Read the whole file at its exact version.

How to install

The exact contents, by fingerprint
mdr add s0mbrad/opentcm_discovery/open-tcm-discovery@sha256:684a6254583b1532

The line at the top of this page pins the label git:20261018.34f43f3, which is what most people want. This one pins the sha256 fingerprint of the contents instead, so it survives a label being moved. Either way the exact version is written to mdr.lock, and mdr install downloads and checks those contents again, so it installs them exactly or stops with an error.

Badge

mdr badge

[![mdr](https://markdownregistry.com/badge/art_obapuykefoneujt7.svg)](https://markdownregistry.com/a/art_obapuykefoneujt7)

2 badge image fetches in 30 days. What the badge shows.

Versions

The registry has recorded 5 versions of this file, between Oct 13, 2026 and Oct 18, 2026, about one every day. Its first recorded version is graded A and its latest B.

versioncommittedcommitsizeaudit
git:20261018.34f43f3 latest2026-10-18 34f43f3 42,336 BB view · diff
git:20261018.64f82242026-10-18 64f8224 35,080 BB view · diff
git:20261013.d0250a02026-10-13 d0250a0 34,870 BB view · diff
git:20261013.9cb006e2026-10-13 9cb006e 31,724 BB view · diff
git:20261013.ac618b72026-10-13 ac618b7 25,597 BA view

Audit of the latest version

B  16 of 17 checks passed. Fixed automatic checks with no AI involved, so the same checks give the same answer every run.
  • fail: No link to a raw IP address (matched: http://47.92.70.12)
  • pass: Frontmatter block present
  • pass: Frontmatter declares a name
  • pass: Frontmatter declares a description
  • pass: Size between 200 bytes and 200 KB (42336 bytes)
  • pass: No zero-width or bidi control characters
  • pass: No instruction hidden inside an HTML comment
  • pass: No link to an exfiltration or paste host
  • pass: No credential-shaped string
  • pass: No instruction to send local credentials anywhere
  • pass: No text hidden with inline styles
  • pass: No prompt-injection phrasing
  • pass: No curl or wget piped into a shell
  • pass: No recursive delete of root, home or parent
  • pass: No instruction to read or print local credentials
  • pass: No base64 blob over 200 characters
  • pass: No script tag

Starting a new one? Use a template for SKILL.md on mdpantry, which passes every one of these checks and says what to change in it.

Source

GitHub

s0mbrad/opentcm_discovery · 31 stars · license MIT · pushed 2026-10-18 · branch main

API

GET https://markdownregistry.com/api/v1/artifacts/art_obapuykefoneujt7
GET https://markdownregistry.com/api/v1/resolve?ref=s0mbrad/opentcm_discovery/open-tcm-discovery
GET https://markdownregistry.com/api/v1/blob/684a6254583b1532e68b709a95d332f93675909e0f7e372a38cf6e39399f67d8

Your AI assistant does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.

Guides: the SKILL.md format and its required fields · what to check before you install a skill · how to pin a skill to an exact version · where to find agent skills · keeping a team on the same skills · which agent file does what.

Browse by kind, by grade B, or by owner.